Added kwargs for tabular suite + fixed missing kwargs in checks (#1515)

deepchecks/tabular/suites/default_suites.py

@@ -32,7 +32,7 @@
            'model_evaluation', 'full_suite']
 
 
-def single_dataset_integrity() -> Suite:
+def single_dataset_integrity(**kwargs) -> Suite:
     """
     Create a suite that is meant to detect integrity issues within a single dataset (Deprecated) .
 
@@ -44,27 +44,27 @@ def single_dataset_integrity() -> Suite:
         'the single_dataset_integrity suite is deprecated, use the data_integrity suite instead',
         DeprecationWarning
     )
-    return data_integrity()
+    return data_integrity(**kwargs)
 
 
-def data_integrity() -> Suite:
+def data_integrity(**kwargs) -> Suite:
     """Create a suite that is meant to detect integrity issues within a single dataset."""
     return Suite(
         'Data Integrity Suite',
-        IsSingleValue().add_condition_not_single_value(),
-        SpecialCharacters().add_condition_ratio_of_special_characters_not_grater_than(),
-        MixedNulls().add_condition_different_nulls_not_more_than(),
-        MixedDataTypes().add_condition_rare_type_ratio_not_in_range(),
-        StringMismatch().add_condition_no_variants(),
-        DataDuplicates().add_condition_ratio_not_greater_than(),
-        StringLengthOutOfBounds().add_condition_ratio_of_outliers_not_greater_than(),
-        ConflictingLabels().add_condition_ratio_of_conflicting_labels_not_greater_than(),
-        OutlierSampleDetection(),
-        FeatureLabelCorrelation().add_condition_feature_pps_not_greater_than()
+        IsSingleValue(**kwargs).add_condition_not_single_value(),
+        SpecialCharacters(**kwargs).add_condition_ratio_of_special_characters_not_grater_than(),
+        MixedNulls(**kwargs).add_condition_different_nulls_not_more_than(),
+        MixedDataTypes(**kwargs).add_condition_rare_type_ratio_not_in_range(),
+        StringMismatch(**kwargs).add_condition_no_variants(),
+        DataDuplicates(**kwargs).add_condition_ratio_not_greater_than(),
+        StringLengthOutOfBounds(**kwargs).add_condition_ratio_of_outliers_not_greater_than(),
+        ConflictingLabels(**kwargs).add_condition_ratio_of_conflicting_labels_not_greater_than(),
+        OutlierSampleDetection(**kwargs),
+        FeatureLabelCorrelation(**kwargs).add_condition_feature_pps_not_greater_than()
     )
 
 
-def train_test_leakage() -> Suite:
+def train_test_leakage(**kwargs) -> Suite:
     """
     Create a suite that is meant to detect data leakage between the training dataset and the test dataset (Deprecated).
 
@@ -76,57 +76,57 @@ def train_test_leakage() -> Suite:
         'the train_test_leakage suite is deprecated, use the train_test_validation suite instead',
         DeprecationWarning
     )
-    return train_test_validation()
+    return train_test_validation(**kwargs)
 
 
-def train_test_validation() -> Suite:
+def train_test_validation(**kwargs) -> Suite:
     """Create a suite that is meant to validate correctness of train-test split, including integrity, \
     distribution and leakage checks."""
     return Suite(
         'Train Test Validation Suite',
-        DatasetsSizeComparison().add_condition_test_train_size_ratio_not_smaller_than(),
-        NewLabelTrainTest().add_condition_new_labels_not_greater_than(),
-        DominantFrequencyChange().add_condition_ratio_of_change_not_greater_than(),
-        CategoryMismatchTrainTest().add_condition_new_category_ratio_not_greater_than(),
-        StringMismatchComparison().add_condition_no_new_variants(),
-        DateTrainTestLeakageDuplicates().add_condition_leakage_ratio_not_greater_than(),
-        DateTrainTestLeakageOverlap().add_condition_leakage_ratio_not_greater_than(),
-        IndexTrainTestLeakage().add_condition_ratio_not_greater_than(),
-        IdentifierLeakage().add_condition_pps_not_greater_than(),
-        TrainTestSamplesMix().add_condition_duplicates_ratio_not_greater_than(),
-        FeatureLabelCorrelationChange().add_condition_feature_pps_difference_not_greater_than()
+        DatasetsSizeComparison(**kwargs).add_condition_test_train_size_ratio_not_smaller_than(),
+        NewLabelTrainTest(**kwargs).add_condition_new_labels_not_greater_than(),
+        DominantFrequencyChange(**kwargs).add_condition_ratio_of_change_not_greater_than(),
+        CategoryMismatchTrainTest(**kwargs).add_condition_new_category_ratio_not_greater_than(),
+        StringMismatchComparison(**kwargs).add_condition_no_new_variants(),
+        DateTrainTestLeakageDuplicates(**kwargs).add_condition_leakage_ratio_not_greater_than(),
+        DateTrainTestLeakageOverlap(**kwargs).add_condition_leakage_ratio_not_greater_than(),
+        IndexTrainTestLeakage(**kwargs).add_condition_ratio_not_greater_than(),
+        IdentifierLeakage(**kwargs).add_condition_pps_not_greater_than(),
+        TrainTestSamplesMix(**kwargs).add_condition_duplicates_ratio_not_greater_than(),
+        FeatureLabelCorrelationChange(**kwargs).add_condition_feature_pps_difference_not_greater_than()
         .add_condition_feature_pps_in_train_not_greater_than(),
-        TrainTestFeatureDrift().add_condition_drift_score_not_greater_than(),
-        TrainTestLabelDrift().add_condition_drift_score_not_greater_than(),
-        WholeDatasetDrift().add_condition_overall_drift_value_not_greater_than(),
+        TrainTestFeatureDrift(**kwargs).add_condition_drift_score_not_greater_than(),
+        TrainTestLabelDrift(**kwargs).add_condition_drift_score_not_greater_than(),
+        WholeDatasetDrift(**kwargs).add_condition_overall_drift_value_not_greater_than(),
     )
 
 
-def model_evaluation() -> Suite:
+def model_evaluation(**kwargs) -> Suite:
     """Create a suite that is meant to test model performance and overfit."""
     return Suite(
         'Model Evaluation Suite',
-        PerformanceReport().add_condition_train_test_relative_degradation_not_greater_than(),
-        RocReport().add_condition_auc_not_less_than(),
-        ConfusionMatrixReport(),
-        SegmentPerformance(),
-        TrainTestPredictionDrift().add_condition_drift_score_not_greater_than(),
-        SimpleModelComparison().add_condition_gain_not_less_than(),
-        ModelErrorAnalysis().add_condition_segments_performance_relative_difference_not_greater_than(),
-        CalibrationScore(),
-        RegressionSystematicError().add_condition_systematic_error_ratio_to_rmse_not_greater_than(),
-        RegressionErrorDistribution().add_condition_kurtosis_not_less_than(),
-        UnusedFeatures().add_condition_number_of_high_variance_unused_features_not_greater_than(),
-        BoostingOverfit().add_condition_test_score_percent_decline_not_greater_than(),
-        ModelInferenceTime().add_condition_inference_time_is_not_greater_than(),
+        PerformanceReport(**kwargs).add_condition_train_test_relative_degradation_not_greater_than(),
+        RocReport(**kwargs).add_condition_auc_not_less_than(),
+        ConfusionMatrixReport(**kwargs),
+        SegmentPerformance(**kwargs),
+        TrainTestPredictionDrift(**kwargs).add_condition_drift_score_not_greater_than(),
+        SimpleModelComparison(**kwargs).add_condition_gain_not_less_than(),
+        ModelErrorAnalysis(**kwargs).add_condition_segments_performance_relative_difference_not_greater_than(),
+        CalibrationScore(**kwargs),
+        RegressionSystematicError(**kwargs).add_condition_systematic_error_ratio_to_rmse_not_greater_than(),
+        RegressionErrorDistribution(**kwargs).add_condition_kurtosis_not_less_than(),
+        UnusedFeatures(**kwargs).add_condition_number_of_high_variance_unused_features_not_greater_than(),
+        BoostingOverfit(**kwargs).add_condition_test_score_percent_decline_not_greater_than(),
+        ModelInferenceTime(**kwargs).add_condition_inference_time_is_not_greater_than(),
     )
 
 
-def full_suite() -> Suite:
+def full_suite(**kwargs) -> Suite:
     """Create a suite that includes many of the implemented checks, for a quick overview of your model and data."""
     return Suite(
         'Full Suite',
-        model_evaluation(),
-        train_test_validation(),
-        data_integrity(),
+        model_evaluation(**kwargs),
+        train_test_validation(**kwargs),
+        data_integrity(**kwargs),
     )

deepchecks/vision/checks/train_test_validation/similar_image_leakage.py

@@ -54,9 +54,9 @@ def __init__(
             self,
             n_top_show: int = 10,
             hash_size: int = 8,
-            similarity_threshold: float = 0.1
-    ):
-        super().__init__()
+            similarity_threshold: float = 0.1,
+            **kwargs):
+        super().__init__(**kwargs)
         if not (isinstance(n_top_show, int) and (n_top_show >= 0)):
             raise DeepchecksValueError('n_top_show must be a positive integer')
         self.n_top_show = n_top_show

tests/tabular/suites/test_suites.py

@@ -59,7 +59,7 @@ def test_generic_suite(
     diabetes_train, diabetes_test, diabetes_model = diabetes_split_dataset_and_model
     city_train, city_test, city_model = city_arrogance_split_dataset_and_model
     iris_train_single, iris_test_single, iris_model_single= iris_split_dataset_and_model_single_feature
-    suite = suites.full_suite()
+    suite = suites.full_suite(imaginery_kwarg='just to make sure all checks have kwargs in the init')
 
     arguments = (
         dict(train_dataset=iris_train_single, test_dataset=iris_test_single, model=iris_model_single),

tests/vision/base/test_suite.py

@@ -234,7 +234,7 @@ def compute(self, context) -> CheckResult:
 
 
 def test_full_suite_execution_mnist(mnist_dataset_train, mnist_dataset_test, mock_trained_mnist, device):
-    suite = full_suite()
+    suite = full_suite(imaginery_kwarg='just to make sure all checks have kwargs in the init')
     arguments = (
         dict(train_dataset=mnist_dataset_train, test_dataset=mnist_dataset_test,
             model=mock_trained_mnist, device=device),
@@ -250,7 +250,7 @@ def test_full_suite_execution_mnist(mnist_dataset_train, mnist_dataset_test, moc
 
 def test_full_suite_execution_coco(coco_train_visiondata, coco_test_visiondata,
                                    mock_trained_yolov5_object_detection, device):
-    suite = full_suite()
+    suite = full_suite(imaginery_kwarg='just to make sure all checks have kwargs in the init')
     args = {'train_dataset': coco_train_visiondata, 'test_dataset': coco_test_visiondata,
             'model':mock_trained_yolov5_object_detection, 'device': device}
     arguments = (